Basin-filling processes and hydrocarbon source rock prediction of low-exploration degree areas in rift lacustrine basins: a case from the Wenchang Formation in low-exploration degree areas, northern Zhu I Depression, Pearl River Mouth Basin, E China

He-Ming Lin , Ho Liu , Xu-Dong Wng , Xin-Wei Qiu ,Yong-To Ju , Jun Meng , Lei Li

a China National Offshore Corporation-Shenzhen Branch, Shenzhen, 518054, China

b School of Ocean Sciences, China University of Geosciences (Beijing), Beijing, 100083, China

c North China University of Science and Technology, Tangshan, 063210, China

d Xi'an Shiyou University, Xi'an, 710065, China

Abstract Hydrocarbon source rocks, as a main geologic factor of petroliferous systems in a sedimentary basin,play a key role in the accumulation of oil and gas and the formation of hydrocarbon accumulations.This study, which focuses on difficulties in prediction of hydrocarbon source rocks in basins or sags with low exploration degree and insufficient hydrocarbon source rock indicators, taking the Wenchang Formation of northern Zhu I Depression,Pearl River Mouth Basin as an example,proposed a hypothesis of“finding lakes and hydrocarbon source rocks”. Detailed steps include, first, determination of the lacustrine basin boundary according to analysis of seismic foreset facies,determination of the depositional area based on the compilation of strata residual thickness maps, determination of the lacustrine basin shape according to deciphering slope break belt system, determination of the fluctuation of paleo-water depth according to biogeochemical indicators of mature exploration areas, determination of the lacustrine basin scale based on analyses of tectonics intensity and accommodation space,which prove the existence of the lacustrine basin and identify the range of semi deep-deep lake; second, further analyses of tectonopalaeogeomorphology, paleo-provenance,palaeoclimate and paleo-water depth to reconstruct the geologic background of the original basin and semideep-deep lacustrine facies, to determine the distribution of semi-deep/deep lacustrine sediments in combination with studies of logging facies,core facies,seismic facies and sedimentary facies,and to rank the sags’potential of developing hydrocarbon source rocks from controlling factors of source-to-sink system development; third, on the basis of regional sedimentary facies analysis, through identification and assessment of seismic facies types of semi-deep/deep lacustrine basins in mature areas, establishing “hydrocarbon source rock facies” in mature areas to instruct the identification and depicting of hydrocarbon source rocks in semideep/deep lacustrine basins with low exploration degree; fourth, through systematical summary of hydrocarbon-rich geological factors and lower limit index of hydrocarbon formation of the sags already revealed by drilling wells (e.g., sag area, tectonic subsidence amount, accommodation space, provenance characteristic,mudstone thickness,water body environment,sedimentary facies types of hydrocarbon source rocks),in correlation with corresponding indexes of sags with low exploration degree,then the evaluation and sorting of high-quality source rocks in areas with sparsely distributed or no drilling wells can be conducted with multi-factors and multiple dimensions. It is concluded that LF22 sag, HZ10 sag and HZ8 sag are II-order hydrocarbon rich sags; whereas HZS, HZ11 and HZ24 are the III-order hydrocarbon-generating sags.

Keywords Wenchang Formation, Pearl River Mouth Basin, Low-exploration degree areas, Rift lacustrine basins,Reconstruction of original basins,Source-to-sink system,Basin filling process,Hydrocarbon source rock facies

1. Introduction

Hydrocarbon source rock is the important geologic factor of the petroliferous system in a sedimentary basin which serves as one of the key factors deciding the accumulation and formation of oil and gas (e.g.,Bruyevich, 1963; Calvert, 1987; Peters et al., 2006;Al-Atta et al.,2014;Arthur and Cole,2014;Jia et al.,2018). Study of effective hydrocarbon source rock distribution in a basin is important for reducing the exploration risks (Zhu et al., 2019). Oil and gas resources mainly accumulate in continental facies in China. Oil and gas preserved in complicated continental facies have been successfully explored and exploited, which develops the oil and gas geological theory system of continental facies (Jia et al., 2018)and successfully advances the prediction methods and technology of hydrocarbon source rocks as well (e.g.,Zhu et al., 2012; Huang et al., 2017; Wang et al.,2019b; Jing et al., 2019; Zhao et al., 2021). However, it's difficult to predict and evaluate hydrocarbon source rocks using traditional methods in areas with relatively low exploration degree,sparsely distributed drilling wells and insufficient hydrocarbon source rock indexes which restrict its oil and gas exploration.

It's well known that organic enrichment is crucial for evaluating the oil and gas prospect of a basin.Organic enrichment is controlled by multiple factors including, e.g., sedimentary rate (e.g., Müller and Suess, 1979; Katz, 2005), original production rate(Calvert, 1987), paleoclimate (Barron, 1985, 1990;Katz, 2005) etc. For inland lacustrine basins, the lacustrine organic abundance is possibly the result of multiple factors, including lacustrine basin tectonics,lacustrine basin shape and depositional process of organic matter (e.g., Suess, 1973; Demaison and Moore, 1980; Olsen, 1990; Garcˊes et al., 1995;Carroll and Bohacs, 1999, 2001; Salmon et al., 2000;Quan et al., 2017; Wang et al., 2019b).

Zhu I Depression is located in the northeast of the Pearl River Mouth Basin and the thickness of Paleogene and Neogene in the basin reaches over 4, 000 m (Mi et al., 2019). The rifting stage of the Pearl River Mouth Basin is from Eocene to early Oligocene, dominated by continental facies, developing hydrocarbon source rocks of shallow lake-deep lake facies of the Wenchang Formation, and shallow lake facies and swamp facies of Enping Formation(Quan et al.,2017;Wang et al., 2019b). Abundant researches of regional geology and petroleum geology have been conducted by previous workers over the years(Pang et al.,2009;Lin and Shi, 2014; Shi, 2015; Shao et al., 2016; Xia et al., 2018; Ma et al., 2020; Mi et al., 2019; Wang et al., 2019a; Shi et al., 2020) using the exploration experiences of continental rift lacustrine basins which effectively support the continuous development of oil and gas exploration in these basins (Mi et al., 2019).Among which, the wide distribution of semi-deep/deep lacustrine facies hydrocarbon source rocks of Wenchang Formation is one of the key factors which decides the oil and gas accumulation and formation in the Pearl River Mouth Basin (Lin and Shi, 2014; Shi,2015; Mi et al., 2019). However, in many sub-sags located in northern Zhu I Depression (Fig. 1), other sags except southern Lufeng area,(e.g.,eastern LF13,western LF13 and LF15) are characterized by its low exploration degree, low study degree, sparsely distributed drilling wells, less systematic research results, less hydrocarbon source rock samples and geochemistry analysis indexes.The breakthrough of oil and gas exploration in areas with sparsely distributed drilling wells lies in successful discovery of deep lake,semi-deep lake and hydrocarbon source rocks, and conduction of prediction of effective and high-quality hydrocarbon source rocks. This study, therefore, taking the basin as our target, on the basis of geologic studies, aims to identify basin range through multiple angles and by using multiple methods and measures.Sedimentary facies was interpreted and divided with constraints of palaeogeomorphology and palaeoenvironment and deposition of deep-to semi-deep lake related with hydrocarbon source rocks was identified as effective ways to conduct research and prediction of hydrocarbon source rocks of areas with no or sparsely distributed drilling wells.

Fig. 1 Tectonic setting of the Pearl River Mouth Basin, tectonic division of Zhu I Depression (A) and main sag distribution of Wenchang Formation in northern Zhu I Depression (B). Distribution characters of northern sags in Zhu I Depression are from original thickness maps(residual thickness map of Wenchang Formation + erosional thickness of key tectonic change periods).

This study takes Wenchang Formation in northern Zhu I Depression (areas with low-exploration degree),the Pearl River Mouth Basin, as a case. This investigation was carried out using the following steps.(1)The basin distribution range was identified based on original basin analysis; (2) The regional source-to-sink system was depicted and basin filling and evolution was studied based on in-depth analyses of basin tectonic palaeogeomorphology, paleo-provenance,palaeoclimate and paleowater depth; (3) The distribution of deep lake and semi-deep lake facies was determined;(4)Sags having hydrocarbon source rocks development potential were sorted in areas with low exploration degree; (5) Seismic facies types were summarized which represent deep lake to semi-deep lake deposits in areas with mature exploration degree; (6) Hydrocarbon source rock facies was established in combination with hydrocarbon source rock evaluation; (7) Geologic factors (lacustrine basin location,sag tectonic subsidence and accommodation space, filling pattern) were developed in mature exploration areas which provide the basis for hydrocarbon source rock prediction in areas with lowexploration degree; and (8) Evaluation and sorting of high-quality hydrocarbon source rocks were carried out in areas with no or sparsely distributed drilling wells using multiple factors and multiple dimensions,which have been conducted on the basis of systematically summarizing sag area, tectonic subsidence amount, accommodation space, accommodation space/rate of sediment supply(A/S)ratio,provenance conditions, mudstone thickness, water body environment, seismic reflections, sedimentary facies type of hydrocarbon source rocks in hydrocarbon-rich sags testified by explorations and in correlation with corresponding indexes of sags with low exploration degree.

2. Regional geologic setting

Zhu I Depression is located in northeastern Pearl River Mouth Basin,trending approximately NE,with an area of approx. 4 × 104km2, including from west to east Enping, Xijiang, Huizhou, Lufeng and Hanjiang sags respectively(Pang et al.,2009;Shao et al.,2016;Xia et al., 2018; Ma et al., 2020; Wang et al., 2019a;Shi et al.,2020)(Fig.1).This study focuses on Huizhou and Lufeng sags in northern Zhu I Depression including subsags of HF33, HZ5, HZ11, HZ10, HZ8, LF7, LF13E,LF13W, LF15, HZ22, HZ24 and LF22 (Fig. 1).

From bottom to top in the depression respectively develop Eocence Wechang Formation, Oligocence Enping and Zhuhai formations, Miocene Zhujiang,Hanjiang and Yuehai formations, Pliocene Wanshan Formation and a the overlying Quaternary. Cenozoic tectonic evolution of Zhu I Depression is divided into a rifting stage and a depression stage (Fig. 2). From Eocene to early Oligocene it belongs to the rifting stage dominated by continental deposition with semideep to deep lake facies of Wenchang Formation as the main hydrocarbon source rock, and shallow lake and swamp facies of Enping Formation as secondary hydrocarbon source rock(Lin and Shi,2014;Shi,2015;Mi et al.,2019).

Previous researchers have conducted a lot of work over the years mainly involving in regional tectonics, sequence stratigraphy, depositional system and petroleum geology (Pang et al., 2009; Zhu et al., 2012; Lin and Shi, 2014; Shi, 2015; Shao et al., 2016; Xia et al., 2018; Ma et al., 2020;Wang et al., 2019a,b; Mi et al., 2019; Shi et al.,2020; Tang et al., 2020). Based on these studies,with reference to exploration experiences of continental rift lacustrine basins,a hypothesis of“sourceto-sink”evaluation system was utilized (Shi, 2015),and guided by “determination of sag-selection of zone”, “source-controlling for finding oil, exploration of areas rich in developing sags, proceeding into deep water area and promotion of oil and gas simultaneously” which provided important theoretical support for hydrocarbon exploration of Zhu I Depression, the Pearl River Mouth Basin (Mi et al.,2019).

3. Data and methods

Drilling wells are mainly located in LF13E sag,LF13W sag and LF15 sag of southern Lufeng Depression in which many hydrocarbon accumulations have been discovered and were testified by relevant analysis data of hydrocarbon source rocks and other geochemistry data. Data of these sags with relatively mature exploration degree will be compared to sags with low exploration degree for further analyses.In addition,a full coverage of 3-D seismic data (3-D seismic data of areas with low exploration degree were re-collected and processed during this study) provides very important foundation support for analyzing sequence stratigraphy, depositional systems, tectonic filling and original basin, deep-to semi-deep lake seismic facies pattern and prediction of favorable hydrocarbon source rocks in areas with low exploration degree in this investigation.

Sequence stratigraphic analysis is to establish sequence framework, based mainly on 3-D seismic data, in combination with sequence information from drilling wells (e.g., lithology, logging, paleontology).Due to the sparsely distributed drilling wells in the study area,boundaries of sequence and systems tracts are respectively identified more according to the unconformable seismic reflection features of 3-D seismic data (e.g., truncation, onlap), and conformable seismic reflection features (toplap, downlap,parallel-sub-parallel etc.). Among which, tectonic sequences are mostly truncation reflection features,whereas the third-order sequence boundary is dominated by seismic onlap, toplap and parallel-subparallel.

Fig.2 Lithostratigraphy and sequences of Zhu I Depression,the Pearl River Mouth Basin.The lithological data and interpretations are based mainly on a combination of data from Well W9. Ages are based on data obtained from CNOOC Shenzhen Branch. En -1 = Member 1 of the Enping Fm.; En -2 = Member 2 of the Enping Fm.; En -3 = Member 3 of the Enping Fm.; En -4 = Member 4 of the Enping Fm.; W -1 and 2=Members 1 and 2 of the Wenchang Fm.;W-3=Member 3 of the Wenchang Fm.;W-4=Member 4 of the Wenchang Fm.;W-5=Member 5 of the Wenchang Fm.; W-6 = Member 6 of the Wenchang Fm. Wenchang Formation is subdivided into Lower and Upper Wenchang Formation,Lower Wenchang Formation includes Member 6, Member 5 and Member 4; whereas Upper Wenchang Formation includes Member 3 and Member 1—2.

Geochemistry, rock slices and zircon dating data of sandstone clastics were acquired from five drilling wells in sags with mature exploration degree of the northern Zhu I Depression (provided by the CNOOCShenzhen Branch). Provenance analysis is based on analyses of these data (cf. Dickinson, 1985; Bhatia et al., 1986; Jian et al., 2013), in combination with research results of seismic stratigraphy palaeogeomorphology and paleouplifts,and focuses on the provenance distribution of the northern Zhu I Depression, nature of mother rocks and tectonic characters of the provenance area(Shao et al.,2016;Wang et al., 2019a).

In this study,we reconstructed the erosion amount of key tectonic changing periods (late depositional period of Lower Wenchang Formation and late depositional period of Upper Wenchang Formation) by referring to the seismic virtual extrapolation method proposed by Liu et al.(2012,2015a,2015b).Based on the reconstructed erosion thickness, in combination with results of compaction correction and identification of slope break belts, the tectonopalaeogeomorphology of Wenchang depositional period in northern Zhu I Depression was reconstructed.

Slope brake belt types including fault slope break zone and deflection slope break zone(Liu et al.,2006)were identified on the basis of 3-D seismic data and through identification of syndepositional faults and analysis of thickness isoline features. Further subdivisions of multistage slope break zone and slope break system in continental lacustrine rift basins were conducted (Liu et al., 2015a, 2017) according to the distribution and mutual combination relationship of slope break zone which lays a foundation for studying provenance area through slope break belt (basin marginal slope break belt)and deep lake to semi-deep lake range (inner basin slope break belt) (Liu et al.,2015a;2015b).

Detailed analyses of paleo-water depth and palaeoclimate were mainly through lithofacies and geochemistry indexes, including water depth change trend reflected by lithofacies (e.g., Bohacs et al.,2000), water depth indicated by Mn/AI (e.g., Toyoda,1993), oxidation-reduction quality of water body indicated by V/Cr et al. (e.g., Nicholas et al., 2004;Zhang et al.,2016),salinity of water body indicated by Sr/Ba and climate humidity indicated by Rb/Cu, Ti/AI et al. (e.g., Adams et al., 2011; Moradi et al., 2016).Lithofacies and geochemistry test data from a total of 12 drilling wells involved in this investigation are provided by the CNOOC-Shenzhen Branch.

We calculated the subsidence rate of all sags,on the basis of calculation of tectonic subsidence amounts of different periods, using a reference for calculation methods of continental lacustrine rift basins by previous workers (Liu et al., 2020), the accommodation space was calculated as total accommodation space = subsidence amount × depositional catchment area which reflects the scale of sags or basins in study.Accommodation space is one of the important factors of hydrocarbon source rock development in sedimentary basins.

The main depositional systems and typical seismic facies types characterizing semi-deep to deep lake facies were identified through comprehensive analyses of seismic reflection texture features (e.g.,amplitude, frequency, continuity), seismic reflection architecture features (e.g., foreset, parallelsubparallel, wavy, worm-shape, chaotic etc.) and seismic reflection shape (e.g., sheet, wedge, flower shape etc.) (Vail et al., 1977; Vail, 1987), in combination with lithofacies analysis of drilling wells in areas with mature exploration degree which provide a basis for identifying distribution of regional depositional systems, establishment of source-to-sink system and prediction of hydrocarbon source rocks in the study area.

4. Identification of areas with low exploration degree and preliminary selection of prospective sags developing deep lake/semi-deep lake deposits

Analyses of sequence stratigraphy, depositional systems,tectonics and palaeogeomorphology are the leading methods of studying sequence and filling process of sedimentary basins (Li, 1995; Ravnas and Steel, 1998; Lemons and Chan, 1999; Lin et al.,2001; Liu et al., 2006, 2015b). Basic geological study and systematic deciphering of deposition,filling and evolution of a basin is especially critical in solving the difficulty of hydrocarbon source rock prediction in areas with sparsely distributed wells or no wells. The basis of hydrocarbon source rock prediction lies in division and identification of regional tectonic sequences, prediction of hydrocarbon source rock development and distribution lies in the reconstruction of primitive basin feature (e.g., tectonic palaeogeomorphology, tectonic subsidence and evolution, palaeoenvironment change, coupling of source-to-sink system and sedimentary pattern and control), and identification of hydrocarbon source rocks lies in important methods of depositional system analysis and semi-deep- to deep lake seismic facies establishment based on seismic interpretations. The above-mentioned is a long-term systematic work and step-by-step workflow does not promise the expected effects. Actually, identification of the developing location of deep and semideep lake areas is key, and should be carried out firstly targeting at prediction of development and potential of hydrocarbon source rocks in areas with low exploration degree. Therefore, five aspects can be carried out at the beginning of the study: (i)identification of lacustrine basin boundary by seismic foreset facies analysis, (ii) identification of deposition areas by compilation of residual strata thickness maps, (iii) identification of lacustrine basin shape by portraying depositional systems,(iv)identification of paleo-water depth fluctuations by bio-chemistry indexes in areas with mature exploration degree, and(v)identification of lacustrine basin scale by analyses of tectonic intensity and accommodation space using simultaneous interpretations of sequence stratigraphy, tectonics and seismic facies based on 3-D seismic data, in combination with drilling well analysis of oil and gas areas with mature exploration degree and the quick “searching” is aimed at“finding lakes”.

4.1. Identification of lacustrine basin boundary by seismic foreset facies analysis

In continental lacustrine rift basins,seismic foreset facies usually corresponds to fan and delta facies,and foreset developing area belongs to river—lake interaction zone, therefore the identification of foreset developing area can provide evidence for preliminary determination of lacustrine basin boundary.

Fig.3 Planar distribution of seismic facies of LF7 and LF13 sags in Lufeng Depression and division of river—lake interaction zone.River-lake interaction area corresponds to foreset developing area in gentle slope belt of the sag;most boundary faults of the sag possibly have direct controls on lacustrine basin range in steep slope belt of the sag.

Taking Wenchang Formation Member 4 in LF7 and LF13 sags of Lufeng Depression as an example,oblique,chaotic-wedge and chaotic foresets (Fig. 3) were identified through 3-D seismic line interpretations which generally correspond to braid delta of gentle slope belt, off-shore subaqueous fan and fan delta.Along gentle slope belt, braid delta develops in the river mouth area whereas regarding steep slope belt lake boundary is controlled by fault boundaries.Based on classification interpretation of seismic foreset facies, in combination with distribution texture of sags, possible river mouth of Member 4 of Wenchang Formation in LF7 and LF13 sags can thus be directly plotted (Fig. 3).

4.2. Identification of depositional area by compilation of strata residual thickness maps

Reconstruction of basin tectono-paleogeomor phology based on eroded thickness and paleo-water depth restoration is the most objective study in identifying topographic features and depositional catchment area of a basin.In a continental rift basin,due to abundant material source supply,adjacent provenance and depositional areas, palaeogeomorphology distribution of targeted intervals can be approximately reflected by strata residual thickness trend assuming that sedimentary surface of any period is flat although certain error could occur because of the strata lacking of studied interval resulted by later period tectonic movement (e.g., tilted lifting and syndepositional fault induced by differential subsidence).

Regional sequence stratigraphic interpretation and identification of Paleogene Wenchang and Enping formations and systematic palaeogeomorphologic studies predicted the hydrocarbon source rocks in northern Zhu I Depression. The original catchment area of the sedimentary basin can thus be identified through quick compilation of residual thickness maps,based on identification of lacustrine basin boundary by seismic foreset facies analysis and according to the developing tendency of thickness of targeted intervals, whereas the catchment area possibly just corresponds to the developing area of the deep lake/semi-deep lake (Fig. 4A).

4.3. Identification of lacustrine basin shape by portraying slope break systems

Fig.4 Residual thickness map and planar interpretation of slope break belt of Wenchang Formation in HF33 sag(A).Multistage slope break belts developed in steep slope belts and gentle slope belts on seismic profiles(B).In figure A three fault slope break belts controlled by faults developed in northern uplifting area which, according to distribution of slope break belt and topography, can be divided into high uplifting and low uplifting area respectively. Slope break-III controls the northern lacustrine basin boundary of HF33 sag. Southern HF33 sag is gentle slope break belt and multiple slope break belts developed: erosional type and fault type developing from south to north. The inner-basin slope break-III controls the southern lacustrine basin boundary. Location of seismic profile in B is shown in A.

Multistage slope break belts develop in lacustrine rift basins regardless of steep slope break belts or gentle slope break belts.Different types of slope break belt associations are closely related with basin deep basement pattern and regional tectonic stress setting.Slope break belts within a same slope break system have obvious genetic symbiotic association and slope break system formed by association of slope break belts is the geomorphologic reflection of the same tectonic stress setting(Liu et al.,2015a;2015b,2017).These slope break belts therefore, are closely related with basin paleotopographic distribution and are one of the factors identifying sedimentary environment of the basin or sag.In multistage slope break belts,basin margin slope break belt controls the development of provenance area (Liu et al., 2006; 2015a; 2015b),whereas slope break belts in the basin are possibly the boundary between delta-shore-shallow lake-deep lake.

The HF33 sag is bounded by a fault at the north and it shows onlap, steep to the north, and gentle to the south. During the deposition of Wenchang-Enping formations, three obvious fault slope break belts developed along the northern steep slope belt, among which, the III-order slope break belt within the basin controlled the boundary of a deep-to semi-deep lake(Fig. 4). Multistage slope break belt developed in southern gentle slope area as well, but with different types and associations of slope break belts. Basin margin slope break belts controlling the southern provenance area are of erosional type; the II-order slope break belts are fault slope break belts which separate alluvial plain and delta; the III-order slope break belts within the basin are reversed type which control the material deposition of local valleys among the II-order slope break belts and are the boundaries of southern sags as well (Fig. 4).

4.4. Identifciation of paleo-water depth fluctuation by biochemistry index in areas with mature oil and gas exploration degree

Formation of hydrocarbon source rocks depends on the living environment of parental biological materials and good preservation conditions of organic matter which are fundamentally decided by the matching of factors including palaeoenvironment and palaeoclimatology during biological reproduction and burial(paleotemperature, paleoprecipitation, paleosalinity,paleo-oxidation-reducibility, paleoproductivity,paleowater depth etc.) (e.g., Suess, 1973; Müller and Suess, 1979; Demaison and Moore, 1980; Barron,1985, 1990; Calvert, 1987; Olsen, 1990; Garcˊes et al., 1995; Carroll and Bohacs, 1999, 2001; Salmon et al., 2000; Katz, 2005; Quan et al., 2017; Wang et al., 2019b). Study on paleoenvironment and palaeoclimate of original basins, therefore, has significant implications for prediction and evaluation of hydrocarbon source rocks. Sedimentary geochemistry and palaeontology data are one of the most important indexes identifying palaeoenvironment and palaeoclimate. Sedimentary geochemistry index is dominated by major,trace REE and isotopes(e.g., Toyoda,1993;Bohacs et al.,2000;Nicholas et al.,2004;Adams et al., 2011; Zhang et al., 2016; Moradi et al., 2016)and palaeontology index mainly includes animal and plant trace fossils including pollen of sediments,foraminifera, algae, ostracoda (Anderson and Lewis,1992; Muller et al., 2003; Bordon et al., 2009;Salonen et al., 2012). Paleowater depth and paleoenvironment can thus be determined through indexes of palaeobiology and geochemistry of drilling wells in areas with mature exploration degree used for reference and evidence of lake basin prediction in areas with low exploration degree.

Fig. 5 Comprehensive analyses of paleowater depth,palaeoclimate and water body environment of Wenchang Formation,Well 14.Notice that deeper water depth and humid climate of the section usually correspond to high TOC value.Geochemistry data were provided by CNOOC-Shenzhen Branch. Value of lithofacies paleo-water depth analysis represents lithofacies rank (cf. Bohacs et al., 2000). W-1 & 2 = Members 1 & 2 of the Wenchang Fm.;W-3=Member 3 of the Wenchang Fm.;W-4=Member 4 of the Wenchang Fm.;W-5=Member 5 of the Wenchang Fm.

Lithofacies and Mn/Al analyses of the study area indicated that during the deposition of Wenchang Formation, it was a deep water setting as a whole,especially Members 3 and 4 reaching the maximum paleowater depth (Fig. 5). Paleowater depth fluctuation of Wenchang Formation is possibly one very solid evidence proving the development of deep-to semideep lake in the study area.Although no direct drilling well data reflect the evolution of paleowater depth in areas with low exploration degree, the tendency of regional paleowater depth fluctuation can be taken as indirect evidence for paleowater depth prediction in these low-exploration-degree areas.

4.5. Identifciation of lacustrine basin scale by analyses of tectonic activity intensity and accommodation space

Tectonics is the most important factor controlling deposition and strata stacking pattern in inland lacustrine basins with strong tectonic activities (Li,1995; Ravnas and Steel, 1998; Lemons and Chan,1999; Lin et al., 2001; Liu et al., 2006, 2015b, 2020),e.g. rift lacustrine basins in eastern China. Tectonic activity controls distribution of the provenance area,changes variation of depositional topography and geomorphology, and significantly influences increase/decrease of accommodation space,fluctuation of lake level and even local climate as well(e.g.,Katz and Liu,1998;Ravnas and Steel,1998;Lemons and Chan,1999;Lin et al., 2001; Liu et al., 2006, 2015a, 2017, 2020).

Tectonic activity intensity during the Eocene-early Oligocene stretching-rifting stage has important controls on the formation and distribution of Paleogene hydrocarbon source rocks in Zhu I Depression. Higher fault activity rate and subsidence rate is favorable for the formation of accommodation space and deep lacustrine basins,insufficient sediment supply tends to form undercompensated sedimentary basin which is favorable for development of high-quality hydrocarbon source rocks of a semi-deep to deep lake (Shi, 2015).Fault activity in Zhu I Depression shows diverse features in different sags across the sedimentary filling: with strong rift and different activity intensity of boundary faults in the sag;the boundary fault controlling the sag with intense activity usually forms a deep rift basin which developed medium-deep lacustrine deposits,e.g.,LF 13,HZ26,FY4 and EP17 sags which are of huge potential(Zhu et al.,2012;Lin and Shi,2014;Shi,2015;Wang et al.,2019b;Mi et al.,2019).

Accommodation space is decided by the difference value of base level and original depositional surface(geomorphology) (Wheeler, 1964). However, it's difficult to restore the accommodation space of ancient strata in basic geological study due to the difficulty in calculating base level and lake level in practical studies(Muto and Steel,2000).Accommodation space can be identified by thickness of preserved sediments in depositional areas of a basin(Muto and Steel,2000),but obviously it is controlled by varied subsidence,tectonic uplifting or variations of base level(e.g.,sea or lake level)(Lemons and Chan,1999).In continental rift basins,however,tectonic uplifting may have more controls on accommodation space(Katz and Liu,1998;Ravnas and Steel, 1998). Based on this, Liu et al.(2020)cited tectonic subsidence data to indicate relative change amplitude and tendency of accommodation space in their quantitative study of tectonic activity intensity and source-to-sink system of Paleogene of inland lacustrine basin-Huanghekou sag,Bohai Bay Basin.The total accommodation space of the basin area was calculated using the formula of: tectonic subsidence amount × depositional catchment area.Therefore,calculations of tectonic activity intensity of different sags or basins can indicate the accommodation size: a larger accommodation space with similar regional palaeoclimate conditions is more favorable for the formation of a deep-to semi-deep lake(Meyers and Ishiwatari, 1993).

Six proved mature sags in Zhu I Depression, the Pearl River Mouth Basin and 8 sags in Hui-Lu area with low exploration degree were chosen for this study and through calculations of tectonic subsidence amount and accommodation space (Liu et al., 2020), the accommodation space distribution of different sags of Wenchang-Enping formations was acquired (Table 1).Relationship between tectonic subsidence amount and accommodation space indicated that tectonics is the most important controlling factor of basin deposition accommodation space which decides both the developing scale of a basin and possibly provides potential accommodation space for deposition of a deep-to semi-deep lake in the basin (Fig. 6). In addition,through calculations of tectonic subsidence amount and accommodation space of areas with mature exploration degree, the lower limits of tectonic subsidence and accommodation space of northern Zhu I Depression can be further acquired, respectively as,the lower limit of tectonic subsidence as 450—550 m and 90—110 km3as the accommodation space(Fig.7).

4.6. Preliminary selection and sorting of potential sags developing deep/semi-deep lake of Wenchang Formation in areas with low exploration degree

Analyses including the above-mentioned five aspects targeted at finding lakes should be conducted successively from multi-aspects including 1) sequence stratigraphic interpretation of regional and key sags providing isochronic stratigraphic framework for seismic facies analysis and thickness map compilation; 2) analysis ofgeometry seismic facies and seismic foreset of 3-D seismic data providing basis for basin depositional system building and more importantly for identifying the fluvial—lake interaction zone of the study area; 3)compilation of residual thickness map of the III-order sequence providing guidance for furthermore identifying sag zones and lacustrine depositional areas in the study area; 4) estimation of seismic foreset type and slope gradient.Seismic foreset type and slope gradient are helpful for judging the provenance character and relative changes of depositional gradient and paleowater depth which therefore can be taken as indirect evidence for judging whether deep-to semi-deep lake facies develop in the sagging areas of the study area.For example, chaotic foreset texture indicated that the study area is located at steep slope belt and possibly associated with deep-to semi-deep lake facies.And“S”type foreset(having complete topset,foreset and bottomset) in which the development of topset indicated that water level of the area increased at that time,accommodation space increased so that terrigenous materials accumulated vertically;development of bottomset indicated that abundant materials deposited in front of the deposits.Whereas according to sedimentary differentiation theory, coarse clastic materials should discharge at topset and foreset locations and areas corresponding to bottomsets usually deposit finegrained sediments. Therefore development of bottomset can be taken as fine-grained terrigenous materials with abundant argillaceous sediments; 5)analysis of tectonic subsidence and tectonic activity intensity mainly aimed at analyzing the accommodation space scale being used to judge the scale of sag deposition area and development potential of prospective hydrocarbon source rocks; 6) paleobiology, geochemistry index and analysis of paleo-water indexing implication aimed at discussing palaeoclimate change,paleo-water depth fluctuation and water body environment from regional point of view and providing reference for areas with low exploration degree.

Table 1 Tectonic activity intensity and accommodation space statistics of Wenchang Formation of the main sags with mature exploration degree in Zhu I Depression and sagging areas with low exploration degree in northern Zhu I Depression.

Fig.6 Relationship between tectonic subsidence amount and accommodation space.R2 of tectonic subsidence amount and accommodation space reaching 0.8 reflects a close relationship between them indicating furthermore the leading role of tectonic activities in rift basins.

Fig.7 Histograms of tectonic subsidence of areas with mature exploration degree in Zhu I Depression(A,B)and areas with low exploration degree in northern Zhu I Depression (C, D).

Through the systematic analyses within a basin range,in combination with the lower limits of tectonic subsidence and accommodation space in areas with mature exploration degree, preliminary selection and sorting was conducted regarding the potential of developing deep-to semi-deep lake in Wenchang Formation in areas with low exploration degree in northern Zhu I Depression, the Pearl River Mouth Basin.Among them, in the Lower Wenchang Formation,except for sags HZ5 and HZ11 having smaller tectonic subsidence amount and accommodation space, other sags are of larger scale. Potentially favorable sags include HF33,HZ24,LF22,LF7,HZ10 and HZ8.Tectonic subsidence amount and accommodation space of the Upper Wenchang Formation are overall smaller, but sags including HF33,HZ24, LF22 and HZ8 are still with development potential of deep-to semi-deep lake.

5. Original basin reconstruction, dynamic background analysis of source-to-sink system and re-selection of favorable sags

Section four of this study focuses on quick searching analysis of“lacustrine basins”aimed at discovering basins which can provide sediments for deep/semideep lakes in combination with basic researches of a basin. However, whether deep-to semi-deep lake deposits really develop in these selected “lacustrine basins (sags) and can be, furthermore, taken as areas developing hydrocarbon source rocks needs comprehensive studies. These studies include in-depth analyzing the controlling factors of deep/semi-deep lake development and original basin features based on exquisite illustration of basin tectonic palaeogeomorphology, palaeo-provenance, palaeo-valleys,palaeoclimate and depositional systems from the integrity of source-to-sink system of sedimentary basins, revealing key geologic factors influencing hydrocarbon source rocks development, which provide evidence for prediction and selection of sags developing favorable hydrocarbon source rocks.

5.1. Tectonopalaeogeomorphology

Tectonopalaeogeomorphology reconstruction belongs to study of original basins which is not only the basis of researches on sequence-depositional system filling process and dynamic setting of a basin,but also provides more scientific and direct evidence for real prediction of distribution of depositional center, sag and hydrocarbon source rocks(Liu et al.,2016,2020).Paleotectonic or palaeogeomorphologic control of a sedimentary basin on hydrocarbon source rocks is a very active hot spot in domestic and overseas researches of sedimentary geology and petroliferous basins (Zhu et al., 2016). Therefore, more and more attentions are paid to reconstruction of eroded thickness, palaeo-water depth-compaction correction and reconstruction of basin tectonic palaeogeomorphology(Liu et al., 2016, 2020).

In this study, regional palaeogeomorphologic reconstruction of Wenchang Formation, northern Zhu I Depression has been conducted based on restoration of eroded thickness in combination with compaction correction (Fig. 8). Overall, the paleotectonics and palaeogeomorphology during the deposition of Wenchang Formation is characterized by (1) uplifts alternating with sags indicating obvious features of rift basins; (2) sag areas are greatly different controlled by existing base and tectonic stress field with overall larger sag area in southern areas (e.g.LF22,LF15 and LF13 sags etc.)than in northern areas(e.g. HZ11, HZ5, HZ10 sags etc.); (3) developing multistage slope break belt among which steep slope belts of the sags are dominated by fault slope break belt and gentle slope belts are dominated by flexure slope break belt; (4) uplifting eroded area can be divided into higher uplift and lower uplift areas (Liu et al., 2016); higher uplift area, as the stable provenance, is the superimposed area of multiple tectonic unconformities including Tg, T83, T80 etc.which is mainly distributed in the northern steepfault zone, Dongsha Uplift and higher location of Huidong low uplift; lower uplift area is controlled by a single tectonic unconformity which is usually a triangle zone of strata overlap-truncation (Liu et al.,2016) influenced by lake level rise and fall and usually is a dynamic provenance area (the area of combined erosion and deposition) and distributed mainly in the northern fault terrace zone; areas between Iorder and II-order slope break belt of Dongsha Uplift and Huidong low uplift and Huilu low uplift is dominated by lower uplift areas.

5.2. Palaeo-provenance and palaeo-valley

Fig. 8 Palaeogeomorphologic map of Lower (A) and Upper (B) Wenchang Formation in northern Zhu I Depression.

Provenance analysis, which plays an important role in determining location and character of sediment provenance, sediment transportation passage and even the deposition of the entire basin and basin tectonic setting (Shao et al., 2016; Wang et al.,2019a; Tang et al., 2020), is an important part of researches on basin and orogenic belt, mainly including studies of sediment compositions,geochemistry features and association feature variation and time-spatial changes of sediment composition. Provenance analysis is especially significant for petroliferous basins, because the location, distance, feature and range of provenance directly influence the formation, transportation and preservation of oil and gas in a basin and are the basis for prediction and evaluation of reservoir and oil and gas exploration (Liu et al., 2017, 2020).

Fig. 9 Provenance comprehensive analysis of Wenchang Formation in northern Zhu I Depression, the Pearl River Mouth Basin. A) Type and distribution of base mother rock of main uplifts (drilling well-seismic joint interpretation); B) Provenance identification according to major and trace element concentration of samples of Well W12. Major and trace elements data were provided by CNOOC-Shenzhen Branch; C)Tectonic projection map of trace elements of Wenchang and Enping formations(drilling well data including W4,W7,W8,W12 and W17 etc.)indicating detrital materials of Wenchang Formation were from continental island arc. Trace elements data were provided by CNOOCShenzhan Branch; D) Lithologic analysis statistics of thin sections and Dickinson sandstone matrix material analysis from Well W17 indicating that Wenchang Formation is mainly a near-source deposition.Identification of thin sections was provided by CNOOC-Shenzhen Branch.

Comprehensive methods were used in this study including seismic stratigraphy, lithology, zircon chronology, geochemistry and micro thin section analysis and provenance distribution, character and tectonic feature analyses were conducted(cf.Dickinson,1985;Bhatia et al., 1986; Jian et al., 2013) providing the basis for establishment of the original basin and source-to-sink system in the study area.It is indicated that uplift base of the study area is dominated by igneous rock, and secondly by sedimentary rock.Wenchang and Enping formations are dominated by felsic igneous rock and quartzose sedimentary rock and sediment provenance of Wechang Formation includes intermediate-basic igneous rock whereas Enping Formation includes granite and more recirculated ancient sedimentary compositions. Sediments of Wenchang Foramtion include more feldspar and debris particles,mainly as near-source deposits. Enping Formation includes more quartz granules and more volcanic detritus, mainly as near-source deposits, which were influenced by peripheral volcanic provenance area.

Meanwhile the distribution of paleo-valleys in uplifting areas was reconstructed (Fig. 10) through seismic interpretation of paleo-valley of base boundary (Tg) in northern Zhu I Depression. Paleo-valley interpretation results indicated that, in total, 74 ancient river systems developed in the study area. In all sags paleo-valleys in long-axis direction are of larger scale and paleo-valleys along the steep slope belt had smaller scale compared to those along gentle slope belt. Paleo-valley is the important passage of depositional systems in a basin(e.g.all types of delta)whereas excessive and overlarge scale drainage of paleo-valleys are not favorable for development of hydrocarbon source rocks. For example, even though HF33 sag is of very large scale(Table 1,Fig.7)however the more numerous paleo-valleys with generally larger scale(Fig.10)possibly restricted the development and distribution of a deep/semi-deep lake. Corresponding discussions will be made later in this article related with developing potential of hydrocarbon source rocks in HF33 sag.

5.3. Palaeoclimate and palaeo-water depth

Fig. 10 Planar distribution of paleo-valleys in main uplifting areas in northern Zhu I Depression. Provenance of paleo-uplifting areas was subdivided into multiple secondary geomorphologic units through present base tectonic features and distribution of main drainage,watershed and ridge lines.

Relationships between regional and local palaeoclimate change and deposition records were investigated through analyses of paleontology and geochemistry elements. Qualitative analysis was conducted on variation tendency of palaeo-water depth,influence of palaeoclimate change on sedimentary sequence, lake water environment and lake level fluctuation was indicated and relationship between palaeo-water depth fluctuation and lake facies deposition was discussed.

Detailed analysis of multiple index reconstruction of palaeo-water depth (lithofacies, major and trace elements), palaeoenvironment (major and trace elements, organic geochemistry) and palaeoclimate was conducted selecting all drilling wells penetrating Paleogene and with biochemistry index in southern Lufeng area in this study. The results indicated that during the deposition of Wenchang to Enping formations the palaeoenvironment was dominated by fresh water, increase of paleo-humidity, decrease of paleosalinity, shoaling of palaeo-water depth and palaeowater depth was positively related with TOC content(Fig.5).However,the palaeoclimate was not obviously related with water depth during the deposition of Wenchang Formation (Fig. 5) which possibly, furthermore, evidenced that tectonics was the dominant control on basin base level,accommodation space and deposition filling in intracontinental rift basins with intense tectonic movements.

5.4. Analysis of regional depositional systems

Building of sedimentary facies and depositional systems of each formation and member of the Paleogene was conducted through analyses of comprehensive seismic facies(including geometry seismic facies,seismic sedimentology),litho(core) facies (core,well logging lithofacies) logging facies (typical sedimentology of electrical logging curves)characteristics and sedimentary environment mark. Fan delta, braid fluvial delta,beach bar,shore-shallow lake,deep lake and characteristic lithologic bodies were divided(Table 2, Fig. 11) and shore-shallow lake, deep-semideep lake will be the key target for study at the next stage.

5.5. Division of source-to-sink system,analysis of depositional system control and reselection of favorable sags

Source-to-sink system refers to the process that the source of sediments formed in the erosion area including weathered and exfoliated granular deposits and dissolved materials are transported to the deposition area or catchment basin and finally deposited there (e.g., Meade, 1982; Leeder, 1997; Anthony and Julian, 1999; Allen, 2008; Sømme et al., 2009, 2013;Sømme and Jackson, 2013; Liu et al., 2019, 2020).Geologic information preserved in source-to-sink system is the record of the entire dynamic process from mountain to basin and the product of interaction of deep lithosphere dynamic process and physical,chemistry and biology and climate conditions of the Earth's surface. The dynamic process and evolution of the Earth's surface can be understood completely only if the formation, transportation and deposition of source of sediments is taken as a whole process for study (e.g., Anthony and Julian, 1999; Allen, 2008;Sømme et al., 2009; Martinsen et al., 2010). Therefore, studies of the source-to-sink system is greatly significant for understanding the evolution and filling process of different sags and the entire basin which will definitely lay a foundation for difference comparison of deposition filling process control and sedimentary response in different sags.

Table 2 Comprehensive comparison of geologic factors for hydrocarbon source rock development in areas with mature exploration degree and areas with low exploration degree in northern Zhu I Depression.Number 1,2 and 3 represents good, moderate and poor developing conditions of hydrocarbon source rocks.

Fig. 11 Planar distribution features of tectonopalaeogeomorphology and depositional systems of Member 4 of Wenchang Formation in northern Zhu I Depression.

The source-to-sink system and its unit of Wenchang Formation in northern Zhu I Depression is subdivided(Fig. 12) based on analyses of tectonopalaeogeomorphology, paleo-provenance and depositional systems and according to factors including source, sediment route,distribution texture of deposition area and sag.In total 10 source-to-sink systems were divided according to source-to-sink associations. We established depositional system development patterns controlled by multiple source-to-sink systems from controlling factors including tectonopalaeogeomorphology (controlling distribution of provenance area and sag),tectonic intensity (controlling sag accommodation space and sediments developing scale), ancient provenance (provenance intensity and sediment body scale), palaeoclimate (regional palaeoclimate controlling lake level fluctuation,water body environment and source sediment supply) and basin texture and filling style (influence of basin undercompensation,isostatic compensation and overcompensation on sediments reflected by A/S ratio). Developing the potential of a deep/semi-deep lake and hydrocarbon source rocks in different sags with low exploration degree were compared and analyzed as well.In total,five source-to-sink depositional patterns developed during the deposition of Lower Wenchang Formation(Fig. 13). Pattern 1 (fan delta-semi-deep depositional system association of small-scale source of sediments supply-high subsidence and undercompensation basin)and pattern 2 (braid fluvial delta-semi-deep depositional system association of medium-scale source of sediments supply-high subsidence-undercompensation basin)are favorable source-to-sink system association patterns for developing a deep/semi-deep lake and hydrocarbon source rocks which were mainly distributed along LF22 northern steep slope belt and HZ08,HZ10, HZ14, LF07, HF33 sags; whereas pattern 4 and pattern 5 with depositional system association of multiple source of sediments-low subsidencecompensated basin are unfavorable for deposition of a deep/semi-deep lake. Six patterns of source-to-sink system developed during the deposition of Upper Wenchang Formation (Fig. 14) and pattern 1 and pattern 2 are favorable source-to-sink system association types for developing a deep/semi-deep lake and hydrocarbon source rocks which mainly developed along northern steep slope belt of HZ05, HZ08, HZ10 and LF22, HZ14, LF07, HF33 sags. From the above comprehensive analyses,a re-sorting of favorable sags can be conducted based on the selection work of Section 4. Favorable sags of Lower Wehchang Formation are ranked as LF22＞HZ8＞HZ10＞HF33＞ LF7＞HZ24＞HZ5(HZ11) and LH22＞HZ8＞HZ10＞HF33＞LF7＞HZ5(HZ11) for Upper Wenchang Formation. It is interesting to find from the development of source-tosink depositional patterns and sag distribution that the calculation results of tectonic subsidence and accommodation space of HF33 are similar as LF22,which are much larger than HZ10, HZ8 sags with larger developing potential of hydrocarbon source rocks and high ranking in preliminary selections. However, in HF33 sag, much coarser detrital materials are easy transported to the basin due to the influence of the surrounding multiple provenances and large valleys which leads to the overcompensation of the basin during most of Wenchang Formation depositional period and higher sand-enrichment degree of the sag unfavorable for deep/semi-deep lake deposition. HZ33 sag therefore ranks after HZ10 and HZ8 sags in sorting of favorable sags.

Fig.12 Division of source-to-sink system of Lower Wenchang Formation in northern Zhu I Depression;in total,10 source-to-sink systems.1-Source-to-sink system of Dongsha Uplift steep slope belt;2-Source-to-sink system of Dongsha Uplift gentle slope belt;3-Source-to-sink system in long axis direction of eastern Lufeng low uplift; 4-Source-to-sink system of steep slope belt of central Lufeng low uplift;5-Source-to-sink system of northern gentle slope belt of eastern Lufeng low uplift;6-Source-to-sink system of northern Haifeng steep slope belt;7-Source-tosink system in long axis direction of eastern Northern Uplift;8-Source-to-sink system of steep slope belt of western Northern Uplift;9-Sourceto-sink system of Huilu low uplift; 10-Source-to-sink system of Huidong.

6. Seismic reflection character of deep/semi-deep lake and establishment of“hydrocarbon source rock facies”

Possible geologic conditions and ancient lake micro setting for development and distribution of hydrocarbon source rocks in areas with low exploration degree are determined in the above quick search of “lacustrine basins” and origin pattern analysis of source-tosink system. Once the sags favorable for developing hydrocarbon source rocks are selected from regional basic geologic study,identification of deep/semi-deep lake facies and hydrocarbon source rocks are needed.Especially from the large set of deep/semi-deep lake mudstone uncovered by drilling wells in areas with mature exploration degree, in combination with seismic reflection characters, seismic responses type and model of a regional deep/semi-deep lake can be established. And finally the pattern of “hydrocarbon source rock seismic facies (hydrocarbon source rock facies)” of the entire northern Zhu I Depression is established based on a thorough consideration of hydrocarbon source rock evaluation index of acquired drilling wells, hydrocarbon source rock prediction in areas with low exploration degree is conducted under the constraints of key geologic factors (e.g., sag tectonic subsidence, accommodation space, filling pattern etc.).“Hydrocarbon source rock facies”study therefore includes well-seismic joint identification of deep/semi-deep lake seismic facies type in areas with mature exploration degree, building of “hydrocarbon source rock facies” of mature areas and hydrocarbon source rock prediction of areas with low exploration degree.

Fig. 13 Summary of source-to-sink system developing pattern of Lower Wenchang Formation in areas with low exploration degree in northern Zhu I Depression. Source-to-sink system and planar distribution of depositional systems corresponded to depositional period of Member 4 of Wenchang Formation. In total, 5 source-to-sink patterns and corresponding number is directly marked in depositional system distribution map so that source-to-sink depositional patterns of different sags are more visualized. Pattern 1 is high subsidenceundercompensation type (fan delta-semi-deep lake association), pattern 2 is high subsidence-undercompensation type (braid fluvial deltasemi-deep lake association), pattern 3 is high subsidence-undercompensation (braid fluvial delta-shore-shallow lake association), pattern 4 is low subsidence-isostatic compensation type (braid fluvial delta-shore-shallow lake association) and pattern 5 is low subsidenceovercompensation type (braid fluvial delta—beach bar-shore-shallow lake association).

6.1. Seismic facies type of deep/semi-deep lake in areas with mature exploration degree

Drilling wells penetrating the Paleogene and with thick mudstone in LF13W, LF13E and LF15 sags were selected in this study. In combination with seismic reflection characters,4 types of deep/semi-deep lake seismic facies were summarized and statistics and analysis were conducted regarding the developing interval,mudstone thickness,geomorphic unit,tectonic subsidence and accommodation space size of the sag of each individual seismic facies type (Fig. 15). Mediumlow amplitude, medium-low frequency wavy-chaotic seismic facies, low amplitude, low frequency-blank reflection seismic facies, low amplitude, medium-low frequency, low continuity seismic facies all correspond to large unit of mudstone (60—200 m, average thickness about 120 m)which are mainly distributed in basin-slope geomorphic unit. Tectonic subsidence amount and accommodation space size of sedimentary period represented by all drilling wells are larger than critical values (Table 1) which indicated sufficiently these three seismic facies types uncovered by drilling wells corresponded to sedimentary environment of deep/semi-deep lake and testified the abovementioned reliability of “finding lake” analysis, sorting of accommodation space and sags developing hydrocarbon source rocks. In addition, medium frequency parallel-subparallel seismic facies develop in the study area as well which, according to traditional textbooks or lots of practical researches, represents a geologic body with lower water body energy and vertical aggradation stacking pattern mostly interpreted as a deep/semi-deep lake deposit (Vail et al., 1977; Vail, 1987). Although this seismic facies type was not proved by drilling well data and possibly due to the interpretation ambiguity during transfer process from seismic facies to sedimentary facies, we made comprehensive judgement whether it belongs to a deep/semi-deep lake deposit through parameters like its sag geomorphic setting (e.g., in basin area),tectonic subsidence and accommodation space size.

Fig. 14 Summary of source-to-sink system developing pattern of Upper Wenchang Formation in areas with low exploration degree in northern Zhu I Depression. Source-to-sink system and planar distribution of depositional systems corresponded to depositional period of Member 3 of Wenchang Formation. In total, 6 source-to-sink patterns and corresponding number is directly marked in depositional system distribution map. Pattern 1 is high subsidence-undercompensation (fan delta-semi-deep lake association), pattern 2 is high subsidenceundercompensation (braid fluvial delta-semi-deep lake association), pattern 3 is high subsidence-undercompensation (braid fluvial deltashore-shallow lake association), pattern 4 is low subsidence-isostatic compensation (braid fluvial delta-shore-shallow lake association),pattern 5 is low subsidence-overcompensation(braid fluvial delta—beach bar-shore-shallow lake association)and pattern 5 is low subsidenceovercompensation (braid fluvial delta-shore-shallow lake association).

6.2. Identification of hydrocarbon source rock seismic facies in areas with mature exploration degree

Deep/semi-deep lake facies is one of the identification standards of hydrocarbon source rocks.Although the above-mentioned deep/semi-deep lake seismic facies is evaluated as potential hydrocarbon source rocks under constraints of multiple factors,identification index of hydrocarbon source rocks is still needed in combination analysis and finally to establish“hydrocarbon source rock facies”.Hydrocarbon source rock evaluation index of mudstone in mature sags of LF13W, LF13E, Lf15 etc. (Fig. 16) indicated organic carbon and organic matter of main intervals of Wenchang Formation are generally of larger values with comprehensive evaluation as good hydrocarbon source rocks among which LF15 as very good. Therefore, it's obvious that the above 4 seismic facies types of deep/semi-deep lake have better hydrocarbon generation conditions, especially the first three types proved by drilling wells, that can be taken as the standard for hydrocarbon source rock prediction in areas with low exploration degree.

6.3. Hydrocarbon source rock prediction in areas with low exploration degree

Fig. 15 Drilling well-seismic joint identification of deep/semi-deep lake in areas with mature exploration degree and summary of deep/semi-deep lake seismic facies types. W -3 = Member 3 of Wenchang Fm.; W -4 = Member 4 of Wenchang Fm.; SF = Seismic facies;GB = Basin geomorphic background; TS = Tectonic subsidence; A = Accommodation space; MT = Mudstone thickness.

Hydrocarbon source rock prediction can be conducted in areas with low exploration degree on the basis of analyses of regional depositional system,seismic facies type of deep/semi-deep lake in mature areas,identification of hydrocarbon source rock facies in mature areas in combination with the abovementioned geologic factors including lacustrine basin distribution,sag tectonic subsidence,accommodation space and filling pattern etc.A detailed analysis of the prediction process was conducted taking Upper Wenchang Formation of HF33 sag as a case.

First,distribution of Upper Wenchang Formation in HF33 sag was identified through “finding lake” analyses (Fig. 4A). Meanwhile calculation results of tectonic subsidence amount and accommodation space indicated that Upper Wenchang Formation in HF33 sag are respectively as 950 m, 238 km3which are much larger than the lower limits of mature hydrocarbonenrichment in this area (Table 1). In combination with the palaeoenvironment characters of fresh water,higher paleo-humidity, lower paleo-salinity and larger paleo-water depth of Wenchang Formation reflected by drilling well data, HF33 sag is preliminarily identified as having potential of developing hydrocarbon source rocks. Secondly, based on detailed analysis of original basin and source-to-sink system, HF33 sag is considered to have similar provenance conditions as mature sags in this area (igneous rock + metamorphic rock) (Fig. 9) and ranks top of sag scale in the entire Huilu area (Figs. 6 and 7) which further testified the potential of developing hydrocarbon source rocks in HF33 sag. Then the dynamic depositional pattern of source-to-sink system indicated that braid fluvial delta-semi-deep lake under filling pattern of high subsidence-undercompensation basin (pattern 2,Fig. 14) developed in Upper Wenchang Formation in eastern HF33 sag; whereas braid fluvial delta-shoreshallow lake under-filling pattern of high subsidenceundercompensation basin (pattern 3, Fig. 14) developed in the western HF33 sag which belongs to the favorable sag for developing deep/semi-deep lake.Seismic reflection type of deep/semi-deep lake of HF33 sag is dominated by medium amplitude, low frequency, wavy-sub-parallel reflection between seismic facies type I and type II of deep/semi-deep lake in areas with mature exploration degree possibly developing better deep-semi-deep lake facies. In addition, the stacking pattern seismic foreset body of Wenchang Formation indicated that geometry shape of foreset body of HF33 sag is characterized by riseretreat in early period to fall-transgression in late period. This reflects that A/S of this depositional period is ≥early period＞1 and late period ＜1 indicating that basin filling type of HF 33 sag is weak undercompensation-weak compensation (Carroll and Bohacs, 1999) and favorable for development of hydrocarbon source rocks. In summary the Upper Wencahng Formation in HF33 sag has certain conditions for developing hydrocarbon source rocks and developed deep/semi-deep lake facies hydrocarbon source rocks with better potential.

Fig. 16 Organic matter types (A) and organic matter abundance character (B) of Member 4 of Wenchang Formation in areas with mature exploration degree in northern Zhu I Depression.Organic matter types are mainly type I-II1 kerogen and organic matter abundance indicating mature areas dominated by high-quality hydrocarbon source rocks.Geochemistry data of hydrocarbon accumulations are provided by CNOOCShenzhen Branch.

7. Comparison of multiple factors and multiple dimensions and evaluation of highquality hydrocarbon source rock

Oil and gas exploration practice and petroleum geology study indicate that the exploration degree of each sag is obviously controlled by hydrocarbonenrichment degree (Shi, 2015; Mi et al., 2019).Development of high-quality hydrocarbon source rocks is decided by the maximum residual thickness of a sag in the plane, sag area, tectonic subsidence rate and sedimentary environment of hydrocarbon source rocks(Shi, 2013). Geologic factors and hydrocarbon accumulation geochemistry index of multiple sags were summarized (Shi, 2013) in pre-researches regarding evaluation of hydrocarbon source rock quality and potential of different sags of Zhu I Depression, which provided good instructions for hydrocarbon source rock evaluation in areas with low exploration degree.Therefore, to further identify the conditions of developing hydrocarbon source rocks selected from main areas with sparsely distributed or no drilling wells in northern Zhu I Depression, and to evaluate if high quality hydrocarbon source rock developed, in this study, a multiple-factor and multiple-dimension evaluation and sorting of high-quality hydrocarbon source rock was conducted based on summarizing the area,tectonic subsidence amount, accommodation space,A/S ratio,provenance conditions,mudstone thickness,water body environment, seismic reflection and sedimentary facies of hydrocarbon source rocks of hydrocarbon-enrichment sags in Zhu I Depression in comparison with corresponding indexes of sags with low exploration degree (Table 2).

The research indicates that Wenchang Formation of LF22, HZ10 and HZ8 sags develop high-quality hydrocarbon source rocks with great hydrocarbongeneration potential being identified as I-order hydrocarbon-enrichment sag; LF7 and HF33 sags with higher hydrocarbon-generation potential as II-order hydrocarbon-enrichment sag; Wenchang Formation of HZ5, HZ11 and HZ24 sags with smaller scale and certain hydrocarbon-generation potential as III-order hydrocarbon-enrichment sag.

8. Discussion

1) The necessary conditions of forming hydrocarbon source rocks include original sediments,hydrocarbon-generating organisms and organic matter preserving condition(e.g.,Suess,1973;Demaison and Moore,1980;Barron,1985,1990;Calvert,1987;Olsen,1990; Garcˊes et al., 1995; Salmon et al., 2000; Katz,2005; Quan et al., 2017; Wang et al., 2019b). In continental rift lacustrine basins, distribution of hydrocarbon source rocks depends on a comprehensive action of multiple factors including basin tectonic activity, paleogeomorphology, source of sediments and palaeoclimate (e.g., Suess, 1973; Demaison and Moore, 1980; Olsen, 1990; Garcˊes et al., 1995;Carroll and Bohacs, 1999, 2001; Salmon et al., 2000;Quan et al.,2017;Wang et al.,2019b).This study aims at prediction and evaluation of hydrocarbon source rocks through geologic comprehensive analyses in areas with sparsely distributed wells and low exploration degree. When lithology and geochemistry indexes of drilling wells are not allowed for directly evaluating and predicting hydrocarbon source rocks,one effective way for this work is finding the lake basin, identifying semi-deep/deep lake sediments(fine-grained sediments), testifying hydrocarbon source rock condition and finally predicting and evaluating hydrocarbon source rocks. Therefore, systematic work should be conducted from the following four aspects: (1) finding lake basins developing hydrocarbon source rocks through comprehensive geologic analyses. Determination of the following factors are necessary to prove the existence of the basin: boundary of the lake basin, boundary between deep and semi-deep lake, lake basin scale and lake level fluctuation; (2) The existence of deep/semi-deep lake sediments needs to be confirmed on the basis of determining the lake basin. Therefore reconstruction of original basin feature and geologic setting of deepsemi-deep lake development is especially important and prediction and evaluation of deep/semi-deep lake deposits in areas with low exploration degree according to controlling factors of source-to-sink system in a basin should be conducted through in-depth researches of tectonic palaeogeomorphology, ancient provenance,palaeolicmate and palaeo-water depth in combination with sedimentary facies identification(especially deep/semi-deep lake facies); (3) Once deep/semi-deep lake deposits are proved to develop in areas with low exploration degree,it needs to confirm in a basin if these sedimentary facies are qualified for developing hydrocarbon source rocks. Hence, identification and depicting of hydrocarbon source rocks in areas with low exploration degree is instructed by the identification and summary of sesimic facies of deep/semi-deep lake in mature areas, hydrocarbon source rock evaluation of drilling wells (related with deep/semi-deep mudstone facies) coupling analyses of geologic factors for hydrocarbon-generation of deep/semi-deep lake facies, establishing “hydrocarbon source rock facies” of mature areas; (4) Finally evaluation and sorting of hydrocarbon source rocks constrained by multiple factors and multiple dimensions in sparsely distributed or no drilling well areas is conducted based on the above-mentioned study and through systematically summarizing hydrocarbon-generating geologic factors uncovered by drilling wells in a basin in comparison with corresponding indexes of sags with low exploration degree.

2)Palaeoenvironment and palaeoclimate are one of the important conditions for formation and development of hydrocarbon source rocks (e.g., Suess, 1973;Müller and Suess, 1979; Demaison and Moore, 1980;Barron,1985,1990;Calvert,1987;Olsen,1990;Garcˊes et al., 1995; Carroll and Bohacs, 1999, 2001; Salmon et al., 2000; Katz, 2005; Quan et al., 2017; Wang et al., 2019b). Palaeoenvironment analysis mainly refers to analysis results of closely neighbouring mature sags in areas with low exploration degree due to the lack of drilling wells in targeted sag and these results have certain consulting implication assuming that the palaeoenvironment had regional overall characters.In fact, some factors of palaeoenvironment such as palaeoclimate was not changed significantly in Zhu I Depression,even in the entire Pearl River Mouth Basin,although each sag was obviously separated during the Paleogene and there might be difference in palaeowater depth and water body environment among sags resulted by variations of paleo-uplift hight and topography due to different tectonic activity. However, the regional character and similar paleo-provenance reflected that paleo-water body of different sags in northern Zhu I Depression were possibly with similar characteristics.For example,almost all drilling wells in mature sags in southern Lufeng area testified the same results of palaeoenvironment which excellently proves this point.

3)In practical study,multiple-factor and multipledimension comparison is proved of high feasibility and credibility compared to evaluation results of highquality hydrocarbon source rocks in areas with low exploration degree. For example, quantity of hydrocarbon generation of Wenchang Formation in southern gentle slope belt LF22 sag was proved to be about 25.5 bt, and HF32-6 was considered to have higher hydrocarbon-generation potential in target evaluation of HF33 sag. New discovery of LF7 sag indicated that abundant hydrocarbon source rocks developed in Lower Wenchang Formation of this sag with a total hydrocarbon-generating amount about 13.4 bt. HZ5-1 was reported as a target of HZ5 sag and the hydrocarbon potential of Enping Formation will be further proved once new discoveries are acquired based on opening up the prospect of exploration in northern Lufeng area.

9. Conclusions

1) It has been a difficulty for hydrocarbon source rock prediction in basins or sags with lower exploration degree, where drilling wells are sparsely distributed,hydrocarbon source rock index is insufficient.A technical system outline of finding lake and finding hydrocarbon source rocks is proposed in this study and the qualitative/semi-quantitative comparisons of multiple-factors and multipledimensions and prediction of high-quality hydrocarbon source rock were conducted.

2) Aiming at development of deep/semi-deep lacustrine basins, through reconstruction of original basin feature and geologic setting of deep/semideep lake development was made through indepth analyses of tectonopalaeogeomorphology,ancient provenance, palaeoclimate and palaeowater depth. In combination with comprehensive analysis of sedimentary facies, deep/semi-deep lake distribution is determined, and sags with hydrocarbon developing potential is sorted according to controlling factors of source-to-sink system development. Favorable sags of Lower Wenchang Formation include LF22＞HZ8＞HZ10＞HF33＞ LF7＞HZ24＞HZ5(HZ1)and Upper Wenchang Formation as LH22＞HZ8＞HZ10＞HF33＞LF7＞HZ5(HZ11).

3) “Hydrocarbon source rock facies” is built for areas with mature exploration degree to instruct hydrocarbon source rock prediction of areas including HF33 sags. This is established on the following aspects: identification of 4 types of seismic facies of deep/semi-deep lake in mature areas(medium-low amplitude, medium-low frequency wavy-chaotic seismic facies, low amplitude, low frequencyblank reflection seismic facies, low amplitude,medium-low frequency low continuity seismic facies), evaluation of hydrocarbon source rocks of drilling wells (related with deep/semi-deep lake mudstone facies)and coupling analyses of geologic factors for hydrocarbon generation of deep/semideep lake (distribution of lake basins, sag tectonic subsidence, accommodation space, filling pattern etc. Upper Wenchang Formation of HZ33 sag is of higher hydrocarbon-generation potential with favorable conditions for developing hydrocarbon source rocks.

4) A systematic summary of hydrocarbon-enrichment geologic factors reflected by drilling wells in the basin (qualitative and quantitative geologic factors, e.g., sag area, tectonic subsidence amount,accommodation space, A/S ratio, provenance conditions, mudstone thickness, water body environment, seismic reflection and sedimentary facies type of hydrocarbon source rocks) is conducted and compared to corresponding indexes of sags with low exploration degree. It is concluded that LF22, HZ10 and HZ8 sags are I-order hydrocarbon-enrichment-generating sag,LF7,HF33 sags are II-order and HZ5, HZ11 and HZ24 are IIIorder hydrocarbon-generating sags.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors thank the Editors and referees Prof.Santanu Banerjee and Prof.Yin-Ye Wu for their helpful comments which improve the quality of the original version. This study was supported by the National Science Foundation of China(Grant No. 41676050).